[PATCH 2/3] Implicit address space conversion on POD assign

[Alp Toker]

On 19/05/2014 07:17, Ronan KERYELL wrote:
>>>>>> On Sun, 18 May 2014 20:00:09 -0700, Richard Smith <richard at metafoo.co.uk> said:
>      Alp> Do we want an extension diagnostic here? As David Chisnall
>      Alp> pointed out
>      >> there are other possible interpretations even if this happens to
>      >> be the one we already use in C mode.
>
>      Richard> Do we have some underlying standard of which this is an
>      Richard> extension?
>
> To have a C++ extension to OpenCL? At least there is the provisional
> specification of http://www.khronos.org/opencl/sycl/
>
> This is something I am starting to play with, but nothing with Clang
> yet: https://github.com/amd/triSYCL
>
> So yes, go ahead to allow OpenCL in a C++ mode directly in the mainstream Clang. :-)

So if Adam's C++ OpenCL language mode gets added to clang you can 
implement the rest of that spec as a library on GitHub without further 
changes to clang?

I like the sound of that :-)

Alp.

-- 
http://www.nuanti.com
the browser experts